SHORT-TERM EVOLUTION IN THE SIZE AND SHAPE OF PEA APHIDS

Phenotypic evolution in contemporary populations can generally be witnessed only when novel selective forces produce rapid evolution. Examples of conditions that have led to rapid evolution include drastic environmental change, invasion of a new predator, or a host-range expansion. In cyclical parthenogens, however, yearly cycles of phenotypic evolution may occur due to the loss of adaptation during recombination in the sexual phase (genetic slippage), permitting an opportunity to observe adaptive evolutionary change in contemporary populations that are not necessarily subject to new patterns of natural selection. In insect herbivores, comparative studies suggest that morphological features that aid individuals in remaining on the plant or exploiting it as a food source are likely targets for selection. Here, we estimated the genetic variability of morphological traits in a cyclical parthenogen, the pea aphid (Acyrthosiphon pisum), to determine the potential for their evolution and we tested the hypothesis that size and/or shape evolves by clonal selection during one season of parthenogenetic reproduction. Genetic variation in a set of morphological traits was estimated using laboratory-reared descendents of clones collected from a single alfalfa field in May 1988 and April 1989 (henceforth, the "early" collections). In both years, there was significant clonal heritability early in the season both for overall morphology and for several individual aspects of size and shape. Because the course of short-term evolutionary change in the multivariate phenotype is a function of patterns of genetic covariance among characters, genetic correlations between size and 12 shape variables were also estimated for these early collections. A comparison between the mean phenotype of each early collection and that of a corresponding "late" collection made from the same field seven to eight clonal generations later in the same years revealed qualitatively similar changes in the average multivariate morphological phenotypes between the time periods in both years, although the difference was only significant for the 1989 samples. The pattern of genetic correlations that we estimated early in the 1989 season between overall size and various shape variables suggests that the observed short-term evolutionary changes in shape could have been due to natural selection acting only to increase overall size. We tested this hypothesis by estimating selection on size using a separate data set in which both demographic and morphological variables were measured on individuals reared under field conditions. Highly significant regressions of individual relative fitness on size were found for two major fitness components. Thus, it is likely that the evolutionary change in morphology that we observed is attributable to natural selection, possibly acting primarily through body size. A shift back to smaller size between the late 1988 and early 1989 collections from the same field suggests that either a cost of recombination or opposing selective forces during overwintering may produce persistent yearly cycles of morphological evolution in this cyclically parthenogenetic species.